Updated to latest double-conversion library version.

runtime/third_party/double-conversion/src/double-conversion.cc

Index: runtime/third_party/double-conversion/src/double-conversion.cc
diff --git a/runtime/third_party/double-conversion/src/double-conversion.cc b/runtime/third_party/double-conversion/src/double-conversion.cc
index 944ddf798eaef829742ff94e6d6f09731913453f..efc1341db2d289063468b2395be52fa10ce311cf 100644
--- a/runtime/third_party/double-conversion/src/double-conversion.cc
+++ b/runtime/third_party/double-conversion/src/double-conversion.cc
@@ -604,7 +604,6 @@ double StringToDoubleConverter::StringToDouble(
   int significant_digits = 0;
   int insignificant_digits = 0;
   bool nonzero_digit_dropped = false;
-  bool fractional_part = false;
 
   bool sign = false;
 
@@ -748,10 +747,8 @@ double StringToDoubleConverter::StringToDouble(
       }
     }
 
-    // We don't emit a '.', but adjust the exponent instead.
-    fractional_part = true;
-
     // There is a fractional part.
+    // We don't emit a '.', but adjust the exponent instead.
     while (*current >= '0' && *current <= '9') {
       if (significant_digits < kMaxSignificantDigits) {
         ASSERT(buffer_pos < kBufferSize);